Identify and agree the sources of knowledge required to determine expected outcomes. Use multiple sources; compare them and cross-check.
Consequence if ignored or violated
There is no way of knowing whether a tested system behaves correctly or not. Inconsistencies and gaps in knowledge cannot be identified.
- What sources of knowledge should be used to derive expected outcomes to validate observed behaviour?
- Do stakeholders approve the use of these sources as a test oracle?
- What confidence can we have in these sources?
- Do expected outcomes need to be documented prior to test execution?
- What is the derivation/heritage/provenance and reliability of these sources?
- What are the levels of authority or precedence of these sources?
- In case of dispute, who or what source will be the final arbiter?